n = 20

QA

prescan

## `summarise()` has grouped output by 'sub', 'nquestion', 'round_text'. You can override using the `.groups` argument.
## Joining, by = "sub"
## Joining, by = "sub"
Prescan Performance
sub correct1 conf conf_correct2
01 0.750 0.333 0.333
02 0.667 0.417 0.417
03 0.792 0.500 0.458
04 0.667 0.625 0.500
05 0.250 0.500 0.083
06 0.708 0.667 0.583
07 0.708 0.333 0.333
08 0.333 0.333 0.333
09 0.792 0.458 0.458
10 0.792 0.458 0.458
11 0.708 0.458 0.458
12 0.833 0.583 0.583
13 0.583 0.917 0.542
14 0.167 0.458 0.125
15 0.750 0.542 0.458
16 0.750 0.500 0.417
17 0.458 0.708 0.417
18 0.625 0.708 0.542
19 0.792 0.583 0.583
20 0.750 0.625 0.542

1 accuracy < 0.5 are highlighted

2 confidence correct < 0.25 are highlighted

scan

Scan Performance
sub correct1 conf conf_correct2
01 0.925 0.025 0.025
02 0.850 0.750 0.725
03 0.650 0.225 0.175
04 0.950 0.950 0.925
05 0.175 0.300 0.125
06 0.775 0.725 0.675
07 0.750 0.000 0.000
08 0.475 0.725 0.450
09 0.975 0.575 0.575
10 0.875 0.625 0.625
11 0.900 0.800 0.775
12 0.800 0.900 0.750
13 0.900 0.925 0.850
14 0.375 0.700 0.350
15 0.225 0.250 0.175
16 0.300 0.300 0.125
17 0.875 0.550 0.550
18 0.750 0.600 0.600
19 0.975 0.950 0.925
20 0.650 0.475 0.425

1 accuracy < 0.5 are highlighted

2 confidence correct < 0.25 are highlighted

postscan

Mean accuracy per subject
sub m1
01 0.9375
02 0.9375
03 1.0000
04 1.0000
05 0.3125
06 1.0000
07 0.9375
08 1.0000
09 1.0000
10 1.0000
11 1.0000
12 1.0000
13 0.8125
14 0.8125
15 1.0000
16 0.8750
17 1.0000
18 1.0000
19 1.0000
20 1.0000

1 accuracy < 0.5 are highlighted

Excluding subject 05, 15, and 16. Should we exclude sub 08 and 14?

Notice: sub1/3/7 have very low confidence.

Prescan analysis

Participant were instructed to answer the expected destination for 3 times during the route: once at Same, once at Overlapping, and once at non-overlapping. They also indicated their confidence towards the choice (sure vs. unsure).

Figures

## `summarise()` has grouped output by 'round_text'. You can override using the `.groups` argument.

## `summarise()` has grouped output by 'nquestion', 'round'. You can override using the `.groups` argument.

Stats

ANVOA for Accuracy:

## Registered S3 methods overwritten by 'lme4':
##   method                          from
##   cooks.distance.influence.merMod car 
##   influence.merMod                car 
##   dfbeta.influence.merMod         car 
##   dfbetas.influence.merMod        car
## $ANOVA
##            Effect DFn DFd          F            p p<.05         ges
## 1           round   1  19  6.9241983 1.644407e-02     * 0.124975005
## 2       nquestion   1  19 76.0754567 4.536688e-08     * 0.633424084
## 3 round:nquestion   1  19  0.9223301 3.489272e-01       0.008496658

ANVOA for Confidence:

## $ANOVA
##            Effect DFn DFd         F            p p<.05        ges
## 1           round   1  19  3.776025 6.695750e-02       0.05876083
## 2       nquestion   1  19 81.230299 2.731736e-08     * 0.71613992
## 3 round:nquestion   1  19  2.064302 1.670474e-01       0.01029772

ANVOA for high confidence accuracy:

## $ANOVA
##            Effect DFn DFd         F            p p<.05        ges
## 1           round   1  19  9.545830 6.032885e-03     * 0.10377224
## 2       nquestion   1  19 93.861783 8.744699e-09     * 0.76533294
## 3 round:nquestion   1  19  3.352941 8.281384e-02       0.01893261

t-test for mean (overlapping segment):

## 
##  Paired t-test
## 
## data:  sub_plot %>% filter(round == 1 & nquestion == 1) %>% .$m and sub_plot %>% filter(round == 2 & nquestion == 1) %>% .$m
## t = -2.4495, df = 19, p-value = 0.02417
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  -0.27817102 -0.02182898
## sample estimates:
## mean of the differences 
##                   -0.15
## 
##  Paired t-test
## 
## data:  sub_plot %>% filter(round == 1 & nquestion == 2) %>% .$m and sub_plot %>% filter(round == 2 & nquestion == 2) %>% .$m
## t = -1.1888, df = 19, p-value = 0.2492
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  -0.20704748  0.05704748
## sample estimates:
## mean of the differences 
##                  -0.075
## 
##  Paired t-test
## 
## data:  sub_plot %>% filter(round == 1 & nquestion == 3) %>% .$m and sub_plot %>% filter(round == 2 & nquestion == 3) %>% .$m
## t = -2.6659, df = 19, p-value = 0.01527
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  -0.15619784 -0.01880216
## sample estimates:
## mean of the differences 
##                 -0.0875

t-test for confidence (overlapping segment):

## 
##  Paired t-test
## 
## data:  sub_plot %>% filter(round == 1 & nquestion == 1) %>% .$conf and sub_plot %>% filter(round == 2 & nquestion == 1) %>% .$conf
## t = 0.27085, df = 19, p-value = 0.7894
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  -0.08409559  0.10909559
## sample estimates:
## mean of the differences 
##                  0.0125
## 
##  Paired t-test
## 
## data:  sub_plot %>% filter(round == 1 & nquestion == 2) %>% .$conf and sub_plot %>% filter(round == 2 & nquestion == 2) %>% .$conf
## t = -2.4908, df = 19, p-value = 0.02217
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  -0.36806043 -0.03193957
## sample estimates:
## mean of the differences 
##                    -0.2
## 
##  Paired t-test
## 
## data:  sub_plot %>% filter(round == 1 & nquestion == 3) %>% .$conf and sub_plot %>% filter(round == 2 & nquestion == 3) %>% .$conf
## t = -1.3708, df = 19, p-value = 0.1864
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  -0.18951388  0.03951388
## sample estimates:
## mean of the differences 
##                  -0.075

t-test for high confidence accuracy (overlapping segment):

## 
##  Paired t-test
## 
## data:  sub_plot %>% filter(round == 1 & nquestion == 1) %>% .$cor_conf and sub_plot %>% filter(round == 2 & nquestion == 1) %>% .$cor_conf
## t = 0, df = 19, p-value = 1
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  -0.1073699  0.1073699
## sample estimates:
## mean of the differences 
##                       0
## 
##  Paired t-test
## 
## data:  sub_plot %>% filter(round == 1 & nquestion == 2) %>% .$cor_conf and sub_plot %>% filter(round == 2 & nquestion == 2) %>% .$cor_conf
## t = -3.3275, df = 19, p-value = 0.003539
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  -0.3665288 -0.0834712
## sample estimates:
## mean of the differences 
##                  -0.225
## 
##  Paired t-test
## 
## data:  sub_plot %>% filter(round == 1 & nquestion == 3) %>% .$cor_conf and sub_plot %>% filter(round == 2 & nquestion == 3) %>% .$cor_conf
## t = -2.2687, df = 19, p-value = 0.03513
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  -0.216287016 -0.008712984
## sample estimates:
## mean of the differences 
##                 -0.1125

Scan analysis

Accuracy per round:

## `summarise()` has grouped output by 'round'. You can override using the `.groups` argument.

Distribution of picture index:

Grouped in 10:

## `summarise()` has grouped output by 'npic_10', 'route'. You can override using the `.groups` argument.

Grouped in 5:

## `summarise()` has grouped output by 'npic_5', 'route'. You can override using the `.groups` argument.

Every picture:

## `summarise()` has grouped output by 'npic', 'route'. You can override using the `.groups` argument.

Postscan analysis

Average accuracy = 0.9669118

## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.